CN112686677A - Customer qualification evaluation method and device based on combination characteristics and attention mechanism - Google Patents
Customer qualification evaluation method and device based on combination characteristics and attention mechanism Download PDFInfo
- Publication number
- CN112686677A CN112686677A CN202011605039.8A CN202011605039A CN112686677A CN 112686677 A CN112686677 A CN 112686677A CN 202011605039 A CN202011605039 A CN 202011605039A CN 112686677 A CN112686677 A CN 112686677A
- Authority
- CN
- China
- Prior art keywords
- initial
- vector
- qualification
- client
- feature
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Images
Abstract
The application belongs to the technical field of intelligent decision making, and provides a customer qualification evaluation method and device based on combined features and an attention mechanism, computer equipment and a computer readable storage medium. According to the method, the client information corresponding to the client is obtained, the client information is converted into the initial characteristic vector corresponding to the client information according to the client information, the initial characteristic vector corresponding to the client information is obtained, the characteristic transformation is performed on the initial characteristic vector to obtain the target characteristic vector corresponding to the initial characteristic vector, the initial characteristic vector and the target characteristic vector are input into a preset client qualification evaluation model based on an attention system, the client is subjected to qualification evaluation, and a client qualification evaluation result corresponding to the client is obtained.
Description
Technical Field
The present application relates to the field of intelligent decision making technologies, and in particular, to a customer qualification evaluation method and apparatus based on a combination feature and attention mechanism, a computer device, and a computer-readable storage medium.
Background
In actual business, the identification of client qualification exists, so that business can be carried out on the client according to the client qualification. For example, in the financial field, customer qualification assessment is important for the development of financial services. However, due to the insufficiency of data such as incomplete client data corresponding to the client, it is difficult to ensure the reliability of the result of the evaluation performed on the client.
Disclosure of Invention
The application provides a customer qualification assessment method and device based on combination characteristics and an attention mechanism, computer equipment and a computer readable storage medium, which can solve the technical problem of low accuracy of financial customer qualification assessment in the traditional technology.
In a first aspect, the present application provides a customer qualification evaluation method based on combined features and attention mechanism, including: acquiring client information corresponding to a client; acquiring an initial characteristic vector corresponding to the customer information according to the customer information; performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector; and inputting the initial characteristic vector and the target characteristic vector into a preset client qualification evaluation model based on an attention mechanism to evaluate the qualification of the client and obtain a client qualification evaluation result corresponding to the client.
In a second aspect, the present application further provides a customer qualification evaluation device based on combined features and attention mechanism, including: the first acquisition unit is used for acquiring client information corresponding to a client; the second acquisition unit is used for acquiring the initial characteristic vector corresponding to the customer information according to the customer information; the transformation unit is used for carrying out characteristic transformation on the initial characteristic vector to obtain a target characteristic vector corresponding to the initial characteristic vector; and the evaluation unit is used for inputting the initial characteristic vector and the target characteristic vector into a preset client qualification evaluation model based on an attention mechanism so as to evaluate the qualification of the client and obtain a client qualification evaluation result corresponding to the client.
In a third aspect, the present application further provides a computer device, which includes a memory and a processor, where the memory stores a computer program, and the processor implements the steps of the customer qualification evaluation method based on the combination features and the attention mechanism when executing the computer program.
In a fourth aspect, the present application further provides a computer readable storage medium having a computer program stored thereon, which, when executed by a processor, causes the processor to perform the steps of the method for client qualification assessment based on combined features and attention mechanism.
The application provides a customer qualification evaluation method and device based on combined features and an attention mechanism, computer equipment and a computer readable storage medium. The method comprises the steps of obtaining customer information corresponding to a customer, converting the customer information into an initial characteristic vector corresponding to the customer information according to the customer information, obtaining an initial characteristic vector corresponding to the customer information, performing characteristic transformation on the initial characteristic vector to obtain a target characteristic vector corresponding to the initial characteristic vector, combining the initial characteristic vector and the target characteristic vector to obtain combined characteristics corresponding to the customer, so that the richness of the characteristic vector corresponding to the customer is increased, the problem that the customer information of the customer is largely absent or excessively sparse is solved, the initial characteristic vector and the target characteristic vector are input into a preset customer qualification evaluation model based on an attention machine to evaluate the qualification of the customer, and a customer qualification evaluation result corresponding to the customer is obtained, therefore, the client characteristics are effectively increased through the combination characteristics corresponding to the client, the weight of the important characteristics on the evaluation result can be increased through the attention mechanism, the client qualification of the client is evaluated based on the combination of the combination characteristics of the client and the attention mechanism, the accuracy and the reliability of the client qualification evaluation can be improved, particularly in the financial field, the client qualification of the financial client is evaluated through the combination of the combination characteristics of the financial client and the attention mechanism, the accuracy and the reliability of the financial client qualification evaluation can be improved, the financial risk of the financial client is reduced, and the possibility of loss is reduced.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present application, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present application, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic flowchart of a customer qualification evaluation method based on combined features and attention mechanism according to an embodiment of the present application;
FIG. 2 is a schematic view of a first sub-process of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application;
FIG. 3 is a second sub-flowchart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application;
fig. 4 is a third sub-flowchart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application;
fig. 5 is a fourth sub-flowchart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application;
FIG. 6 is a diagram illustrating an exemplary model of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application;
FIG. 7 is a schematic block diagram of a customer qualification evaluation device based on a combination feature and attention mechanism according to an embodiment of the present application; and
fig. 8 is a schematic block diagram of a computer device provided in an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are some, but not all, embodiments of the present application. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
Referring to fig. 1, fig. 1 is a schematic flow chart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application. As shown in FIG. 1, the method includes the following steps S11-S14:
and S11, obtaining the client information corresponding to the client.
Specifically, when the qualification evaluation is performed on the client, although there may be a problem that a large amount of client information of the client is missing or too sparse, this is only a relative problem, and in the actual business development process, a plurality of pieces of client information corresponding to the client are obtained, so that a plurality of pieces of client information corresponding to the client can be obtained. For example, in the financial field, when a financial service is performed on a client, the client qualification evaluation is performed on the client, to judge whether the customer satisfies the condition of developing financial business, the customer information of the customer in various aspects is often needed, so as to comprehensively judge the client and determine whether the client qualification of the client meets the condition for developing the financial service, in the process, although the information of the client has a large amount of missing or too sparse problems, but the financial qualification evaluation of the client can not be carried out by only using the simple client information such as the name or the sex of the client, to judge whether to carry out financial service for the client or to obtain multiple pieces of client information corresponding to the client, the client qualification judgment is carried out on the client, and the client information may have a large amount of missing or too sparse problems compared with the comprehensive and complete analysis of the client qualification corresponding to the client. Therefore, in the financial field, when a financial transaction is performed on a client, a plurality of pieces of client information corresponding to the client, such as the name, sex, age, address, occupation, income, and the like of the client, are acquired.
And S12, acquiring the initial characteristic vector corresponding to the customer information according to the customer information.
Specifically, after obtaining a plurality of pieces of customer information corresponding to customers, each piece of customer information may be converted into a feature vector corresponding to each piece of customer information, for example, for numerical features included in the customer information, the numerical features may be directly converted into corresponding feature vectors, and for non-numerical features included in the customer information, the word vectors may be converted into corresponding feature vectors to obtain each initial feature corresponding to each piece of customer information, so as to obtain an initial feature vector corresponding to the customer information, and obtain an initial feature vector corresponding to the customer information.
And S13, performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector.
Specifically, after an initial feature vector corresponding to the customer information is obtained, feature transformation is performed on the initial feature vector, for example, operations such as pairwise multiplication and triplex multiplication are performed on all the initial feature vectors, and a quadratic power of each initial feature vector or a cubic power of each initial feature vector are obtained, so that feature transformation is performed on the initial feature vectors to obtain new vector features corresponding to the initial feature vectors, and the new vector features are used as target feature vectors to obtain target feature vectors corresponding to the initial feature vectors, so that the feature vectors corresponding to the customers are supplemented, and the problem that a large amount of customer information of the customers is missing or too sparse is solved.
And S14, inputting the initial characteristic vector and the target characteristic vector into a preset client qualification evaluation model based on an attention mechanism to evaluate the qualification of the client and obtain a client qualification evaluation result corresponding to the client.
The basic idea of Attention mechanism (english is Attention) in a computer is to let a system learn Attention and to be able to ignore irrelevant information and focus on key information, and the Attention model corresponding to the Attention mechanism includes an Encoder-Decoder framework, a SoftAttention model, a self Attention model, and the like.
Specifically, a preset client qualification evaluation model is pre-constructed based on an attention mechanism, and the preset client qualification evaluation model is trained by using the existing client information and a client qualification evaluation result corresponding to the client information. After the target characteristic vector corresponding to the initial characteristic vector is obtained, the initial characteristic vector and the target characteristic vector are combined to obtain combined characteristics corresponding to the customer, the combined characteristics can be combined in a splicing mode to obtain a vector matrix (namely a vector queue) formed by combining the initial characteristic vector and the target characteristic vector, the vector matrix is input into a preset customer qualification evaluation model based on an attention mechanism, and qualification evaluation is performed on the vector matrix corresponding to the customer through a pre-trained preset customer qualification evaluation model based on the attention mechanism to obtain a customer qualification evaluation result corresponding to the customer.
In the embodiment of the application, customer information corresponding to a customer is obtained, the customer information is converted into an initial feature vector corresponding to the customer information according to the customer information, an initial feature vector corresponding to the customer information is obtained, feature transformation is performed on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector, the initial feature vector and the target feature vector are combined to obtain combined features corresponding to the customer, so that the richness of the feature vector corresponding to the customer is increased to overcome the problem that the customer information of the customer is largely absent or excessively sparse, the initial feature vector and the target feature vector are input into a preset customer qualification evaluation model based on an attention system to evaluate the qualification of the customer and obtain a customer qualification evaluation result corresponding to the customer, therefore, the client characteristics are effectively increased through the combination characteristics corresponding to the client, the weight of the important characteristics on the evaluation result can be increased through the attention mechanism, the client qualification of the client is evaluated based on the combination of the combination characteristics of the client and the attention mechanism, the accuracy and the reliability of the client qualification evaluation can be improved, particularly in the financial field, the client qualification of the financial client is evaluated through the combination of the combination characteristics of the financial client and the attention mechanism, the accuracy and the reliability of the financial client qualification evaluation can be improved, the financial risk of the financial client is reduced, and the possibility of loss is reduced.
Referring to fig. 2, fig. 2 is a schematic view of a first sub-flow of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application. As shown in fig. 2, in this embodiment, the step of obtaining the initial feature vector corresponding to the customer information according to the customer information includes:
s21, judging whether the customer information is numerical value information;
s22, if the customer information is numerical information, converting the customer information into corresponding initial feature vectors according to a preset numerical conversion mode;
and S23, if the customer information is non-numerical information, converting the customer information into the corresponding initial feature vector according to a preset word vector conversion mode.
Specifically, the client information corresponding to the client can be divided into numerical information and non-numerical information, because the vector itself is a numerical value, if the client information is numerical information, the corresponding initial feature vector can be directly obtained from the numerical information of the client according to a preset numerical value conversion mode, and if the client information is non-numerical information, the non-numerical information can be converted into the feature vector through a Word vector (english is Word2Vec) technology according to a preset Word vector conversion mode, so that the initial feature vector corresponding to the client information is obtained. For example, if the number of the client information is n, all the feature vectors are X1,X2…Xn，X1,X2…XnEach one of which isOne is a corresponding feature vector.
Referring to fig. 3, fig. 3 is a second sub-flowchart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application. As shown in fig. 3, in this embodiment, the step of performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector includes:
s31, acquiring a plurality of initial feature vectors corresponding to the customer information;
s32, sequencing the initial characteristic vectors according to a preset first sequencing mode to obtain a first sequencing queue;
s33, acquiring a first preset number of initial feature vectors contained in the first sequencing queue, and taking the first preset number of initial feature vectors as initial target feature vectors;
and S34, performing feature transformation on the initial target feature vector to obtain a target feature vector corresponding to the initial target feature vector.
Specifically, if the client has a plurality of pieces of client information, each piece of client information has a corresponding initial feature vector, and the client has a plurality of initial feature vectors. In the initial feature vectors, because the different initial feature vectors have different degrees of closeness with the qualification of the assessment client, the initial feature vector which is most relevant to the qualification of the assessment client can be screened from the initial feature vectors, the initial feature vector which is most relevant to the qualification of the assessment client is fully utilized to assess the qualification of the client, all the initial feature vectors are avoided being processed, and the accuracy and the efficiency of assessing the qualification of the client can be improved. The method includes the steps of obtaining all initial feature vectors corresponding to the customer information, wherein the all initial feature vectors correspond to a plurality of initial feature vectors, sorting all the initial feature vectors according to a preset sorting mode according to the closeness degree of all the initial feature vectors and the customer qualification to obtain a first sorting queue corresponding to all the initial feature vectors, obtaining a first preset number of the initial feature vectors contained in the first sorting queue if the preset sorting mode sorts all the initial feature vectors according to the sequence from high closeness degree of all the initial feature vectors and the customer qualification, taking the first preset number of the initial feature vectors as initial target feature vectors, performing feature transformation on the initial target feature vectors to obtain target feature vectors corresponding to the initial target feature vectors, and then through screening out the initial characteristic vector most relevant to the qualification of the assessment client and generating the combination characteristic corresponding to the initial characteristic vector most relevant to the qualification of the assessment client, the qualification of the client is assessed through the initial characteristic vector and the target characteristic vector corresponding to the combination characteristic, and because the target characteristic vector is equivalent to increase the richness of the characteristic vector most relevant to the qualification of the assessment client, the accuracy of the qualification of the client can be improved, and because all the initial characteristic vectors of the user do not need to be assessed, only the characteristic vector most relevant to the qualification of the assessment client is assessed, the efficiency of the qualification of the client can be improved.
Referring to fig. 4, fig. 4 is a third sub-flowchart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application. As shown in fig. 4, in this embodiment, the step of sorting the plurality of initial feature vectors according to a preset first sorting manner to obtain a first sorting queue includes:
s41, obtaining a preset qualification vector corresponding to the client qualification;
s42, calculating respective first correlation of each initial characteristic vector and the preset qualification vector;
s43, calculating the first importance corresponding to each initial feature vector and the preset qualification vector;
s44, sorting the initial eigenvectors according to the first correlation and the first importance corresponding to each initial eigenvector and the preset qualification vector respectively to obtain a first sorting queue.
In particular, to the clientThe method comprises the steps of obtaining a plurality of corresponding initial feature vectors, obtaining preset qualification vectors corresponding to customer qualification, calculating the correlation between each initial feature vector and the preset qualification vectors, calculating the importance of each initial feature vector and the preset qualification vectors, sequencing the initial feature vectors according to the first correlation and the first importance of each initial feature vector and the preset qualification vectors, sequencing the initial feature vectors according to the correlation between the initial feature vectors and the preset qualification vectors to obtain a first correlation sequencing queue, sequencing the first correlation sequencing queue according to the importance of the initial feature vectors and the preset qualification vectors to obtain a first sequencing queue, and sequencing all the initial feature vectors according to the correlation and the importance of the initial feature vectors when evaluating the customer qualification simultaneously And then screening a preset number of initial characteristic vectors from the first sequencing queue to obtain initial target characteristic vectors. For example, in the financial field, for all initial feature vectors corresponding to customer information, the relevance and importance of each initial feature vector and a preset Y value of the qualification assessment of the financial customer are calculated, all the initial feature vectors are sorted to obtain a sorting queue, and then M feature vectors X 'with the most top relevance and importance can be taken out of the first sorting queue'1,X'2…X'm。
Wherein, calculating the first correlation between each initial feature vector and the preset qualification vector, and calculating the feature correlation by using a pearson coefficient, and the calculation method is as follows:
wherein Xi is a single initial feature vector, Yi is a value corresponding to a Y value contained in the preset qualification vector,
and
is the mean of the eigenvectors and the Y values, pX,YThe correlation size (or referred to as correlation height) between the initial feature vector and the preset qualification vector is obtained.
Wherein, calculating the first importance of each initial feature vector and the preset qualification vector, the first importance of the features can be calculated by using random forests. The calculation method for any one initial feature vector X in the random forest is as follows:
for each decision tree in the random forest, calculating its out-of-bag data error, denoted err1, using the corresponding out-of-bag data;
noise interference is randomly added to the initial eigenvector X of all samples of the data outside the bag, and the error of the data outside the bag at this time is calculated and is recorded as err 2.
If the random forest has n trees, then the importance of the initial feature vector X is:
sigma (err1-err2)/n formula (2)
In an embodiment, the step of performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector includes:
and performing n x n operation on all the initial feature vectors to obtain a plurality of first conversion feature vectors, and taking the first conversion feature vectors as target feature vectors, wherein n is a natural number and is used for describing the n initial feature vectors.
Specifically, n × n operation is performed on all the initial feature vectors, for example, every two initial feature vectors in all the initial feature vectors are used as one group, every two groups of initial feature vectors are multiplied, namely, the initial feature vectors are multiplied by each other to obtain a first conversion feature vector, or every three initial feature vectors in all the initial feature vectors are used as one group, every two groups of initial feature vectors are multiplied, namely, the initial feature vector is multiplied by three times to obtain a first conversion feature vector, and the first conversion feature vector is used as a target feature vector, all the first conversion feature vectors can be used as target feature vectors, therefore, the feature vectors corresponding to the clients are supplemented, and the problem that the client information of the clients is largely lost or excessively sparse is solved.
In an embodiment, before the step of using the first converted feature vector as a target feature vector, the method further includes:
performing preset power square operation on the initial characteristic to obtain a second conversion characteristic vector corresponding to the initial characteristic vector;
the step of using the first converted feature vector as a target feature vector comprises:
and taking the first conversion feature vector and the second conversion feature vector as the target feature vector.
Specifically, the initial feature is subjected to a predetermined power operation, for example, each initial feature vector is subjected to a quadratic operation and/or a cubic operation to obtain a second conversion feature vector corresponding to the initial feature vector, the second conversion feature vector is also used as the target feature vector, all the second conversion feature vectors can be also used as the target feature vector, and thus all the first conversion feature vectors and all the second conversion feature vectors are used as the target feature vectors.
For example, in an example, the step of performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector further includes:
multiplying all the initial feature vectors pairwise to obtain a first sub-conversion feature vector;
performing three-three multiplication on all the initial feature vectors to obtain a second sub-conversion feature vector;
performing quadratic operation on each initial feature vector in all the initial feature vectors to obtain a third sub-conversion feature vector;
performing cubic operation on each initial feature vector in all the initial feature vectors to obtain a fourth sub-conversion feature vector;
and taking the first sub-conversion feature vector, the second sub-conversion feature vector, the third sub-conversion feature vector and the fourth sub-conversion feature vector as target feature vectors.
Specifically, if the number of the initial feature vectors is M, the initial feature vector may be X'1,X'2…X'mThe M initial feature vectors are multiplied by two and three times respectively, and quadratic and cubic operations of each initial feature vector are solved to obtain a first sub-conversion feature vector, a second sub-conversion feature vector, a third sub-conversion feature vector and a fourth sub-conversion feature vector of new features corresponding to each initial feature vector as target feature vectors, and all the new features, namely the first sub-conversion feature vector, the second sub-conversion feature vector, the third sub-conversion feature vector and the fourth sub-conversion feature vector, can be used as the target feature vectors, so that new combination features are added, the condition that a decimal result is not too small and a large result is too large is ensured, and the accuracy of customer qualification evaluation can be improved.
Referring to fig. 5, fig. 5 is a fourth sub-flowchart of a customer qualification evaluation method based on combination features and attention mechanism according to an embodiment of the present application. As shown in fig. 5, in this embodiment, the step of using the first converted feature vector and the second converted feature vector as the target feature vector includes:
s51, taking the first conversion feature vector and the second conversion feature vector as conversion feature vectors and sequencing the conversion feature vectors according to a preset second sequencing mode to obtain a second sequencing queue;
and S52, acquiring a second preset number of the conversion feature vectors contained in the second sorting queue, and taking the second preset number of the conversion feature vectors as the target feature vectors.
Specifically, if the first conversion feature vector and the second conversion feature vector correspond to a large number of conversion feature vectors, in a plurality of conversion feature vectors, because different conversion feature vectors have different degrees of closeness with the qualification of the assessment client, the conversion feature vector most relevant to the qualification of the assessment client can be screened out from the plurality of conversion initial feature vectors, the conversion feature vector most relevant to the qualification of the assessment client is fully utilized to assess the qualification of the client, processing of all conversion feature vectors is avoided, and accuracy and efficiency of assessing the qualification of the client can be improved. All the first conversion feature vectors and all the second conversion feature vectors are collectively called as conversion feature vectors by obtaining all the first conversion feature vectors and all the second conversion feature vectors, the conversion feature vectors correspond to a plurality of feature vectors, all the conversion feature vectors are sorted according to a preset second sorting mode according to the closeness degree of all the conversion feature vectors with the qualification of the customer, the second sorting mode can be the same as the first sorting mode to obtain a second sorting queue corresponding to all the conversion feature vectors, if the preset second sorting mode is also sorted according to the closeness degree of all the conversion feature vectors with the qualification of the customer from high to low, a second preset number of the conversion feature vectors contained in the second sorting queue are obtained, and the second preset number of the conversion feature vectors are used as the target feature vectors, therefore, the conversion characteristic vector which is most relevant to the qualification of the evaluated customer is screened out, the conversion characteristic vector and the initial characteristic vector are combined into the combined characteristic, so that the combined characteristic which is most relevant to the qualification of the evaluated customer is generated, the qualification of the customer is evaluated through the initial characteristic vector and the target characteristic vector which correspond to the combined characteristic, the target characteristic vector is equivalent to the increase of the richness of the characteristic vector which is most relevant to the qualification of the evaluated customer, the accuracy of the qualification of the customer can be improved, and the efficiency of the qualification of the customer can be improved as all the conversion characteristic vectors of the customer are not required to be evaluated, and only the conversion characteristic vector which is most relevant to the qualification of the evaluated customer is evaluated.
In an embodiment, the step of taking the first conversion feature vector and the second conversion feature vector as conversion feature vectors and performing sorting according to a preset second sorting manner to obtain a second sorting queue includes:
acquiring the preset qualification vector corresponding to the client qualification;
calculating a second correlation corresponding to each conversion characteristic vector and the preset qualification vector;
calculating a second importance corresponding to each conversion feature vector and the preset qualification vector;
and sequencing the plurality of conversion characteristic vectors according to the first correlation and the second importance respectively corresponding to each conversion characteristic vector and the preset qualification vector to obtain a second sequencing queue.
Specifically, for a conversion feature vector corresponding to the customer and composed of a plurality of first conversion feature vectors and second conversion features, a preset qualification vector corresponding to qualification of the customer is obtained, second relevance corresponding to each of the conversion feature vectors and the preset qualification vector is calculated, second importance corresponding to each of the conversion feature vectors and the preset qualification vector is calculated, the conversion feature vectors are sorted according to the second relevance and the second importance corresponding to each of the conversion feature vectors and the preset qualification vector, the conversion feature vectors can be sorted according to the second relevance of the conversion feature vectors and the preset qualification vector to obtain a second relevance sorting queue, and then the second importance of the conversion feature vectors and the preset qualification vector is calculated, and sequencing the second relevance sequencing queue to obtain a second sequencing queue, sequencing all the conversion characteristic vectors by simultaneously combining the second relevance and the second importance when the conversion characteristic vectors are used for evaluating the client qualification to obtain a second sequencing queue, and screening a second preset number of the conversion characteristic vectors as target characteristic vectors. For example, in the financial field, all conversion characteristics corresponding to customer informationVector, calculating the relevance and importance of each conversion feature vector and the preset Y value of the financial customer qualification assessment, sorting all the conversion feature vectors to obtain a second sorting queue, and then taking out T conversion feature vectors X with the top relevance and importance from the second sorting queue "1,X”2…X”T. The second correlation between each of the transformed feature vectors and the preset qualification vector is calculated, or the second correlation between each of the transformed feature vectors and the preset qualification vector is calculated by using a pearson coefficient, or the second importance of the feature is calculated by using a random forest, and the process of calculating the feature correlation and the importance may refer to the respective calculation processes, which is not described herein again.
Further, M most-significant feature vectors X 'are fetched from the first sorting queue'1,X'2…X'mTaking T conversion feature vectors X with the most advanced correlation and importance from the second sorting queue "1,X”2…X”TFor example, please refer to fig. 6, fig. 6 is an exemplary diagram of a model of a customer qualification evaluation method based on a combination feature and attention mechanism provided in an embodiment of the present application, and as shown in fig. 7, when the attention mechanism is used, an attention model with a model name of Sec, Sec1 represented by feature X 'may be adopted'1,X'2…X'mAnd X "1,X”2…X”TThe matrix is formed by combining the simple characteristic vectors, and the matrix is a new characteristic after being used. Sec2 is a convolutional network that can use 3 × 3 max pooling, with a convolution kernel of 3 × 3 size. Sec3 is a convolutional network, using 3 × 3 max pooling, with a convolution kernel of 7 × 7 size. Sec4 is an upsampling module, where upsampling uses classical linear interpolation, enabling resizing of extracted features through the module to sum Sec1The result of the size agreement is output, and the extracted features are resized by the module to agree with the Sec1 output size. Sec5 multiplies the outputs of Sec1 and Sec4 to obtain a new characteristic, and the new characteristic is sent to Sec6 for qualification evaluation, so that an evaluation result of the qualification of the client is obtained. The model of Sec6 can be selected as a convolutional neural network, linear regression, Xgboost, etc., and AUC evaluation criteria can be used for severely unbalanced samples, and mean square error, cross entropy, AUC can be used for equalized samples.
It should be noted that, the customer qualification assessment method based on the combined feature and attention mechanism described in the above embodiments may recombine the technical features included in different embodiments as needed to obtain a combined implementation, but all of them are within the protection scope claimed in the present application.
Referring to fig. 7, fig. 7 is a schematic block diagram of a customer qualification evaluation device based on combination features and attention mechanism according to an embodiment of the present application. Corresponding to the above customer qualification evaluation method based on the combination feature and the attention mechanism, the embodiment of the present application further provides a customer qualification evaluation device based on the combination feature and the attention mechanism. As shown in fig. 7, the client qualification evaluation device based on combination characteristics and attention mechanism comprises a unit for executing the above-mentioned client qualification evaluation method based on combination characteristics and attention mechanism, and the client qualification evaluation device based on combination characteristics and attention mechanism can be configured in a computer device. Specifically, referring to fig. 7, the client qualification evaluation device 70 based on combination features and attention mechanism includes a first obtaining unit 71, a second obtaining unit 72, a transforming unit 73 and an evaluating unit 74.
The first obtaining unit 71 is configured to obtain client information corresponding to a client;
a second obtaining unit 72, configured to obtain an initial feature vector corresponding to the customer information according to the customer information;
a transforming unit 73, configured to perform feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector;
and the evaluation unit 74 is configured to input the initial feature vector and the target feature vector into a preset customer qualification evaluation model based on an attention mechanism, so as to perform qualification evaluation on the customer and obtain a customer qualification evaluation result corresponding to the customer.
In one embodiment, the second obtaining unit 72 includes:
the judging subunit is used for judging whether the client information is numerical value information;
the first conversion subunit is used for converting the customer information into a corresponding initial feature vector according to a preset numerical value conversion mode if the customer information is numerical value information;
and the second conversion subunit is used for converting the customer information into the corresponding initial characteristic vector according to a preset word vector conversion mode if the customer information is non-numerical information.
In one embodiment, the transforming unit 73 includes:
the first acquisition subunit is used for acquiring a plurality of initial feature vectors corresponding to the customer information;
the first sorting subunit is configured to sort the plurality of initial feature vectors according to a preset first sorting mode to obtain a first sorting queue;
a second obtaining subunit, configured to obtain a first preset number of the initial feature vectors included in the first sorting queue, and use the first preset number of the initial feature vectors as initial target feature vectors;
and the transformation subunit is used for performing characteristic transformation on the initial target characteristic vector to obtain a target characteristic vector corresponding to the initial target characteristic vector.
In one embodiment, the first ordering sub-unit comprises:
a third obtaining subunit, configured to obtain a preset qualification vector corresponding to the client qualification;
the first calculating subunit is configured to calculate a first correlation corresponding to each of the initial feature vectors and the preset qualification vector;
the second calculating subunit is configured to calculate a first importance corresponding to each of the initial feature vectors and the preset qualification vectors;
and the second sorting subunit is used for sorting the plurality of initial characteristic vectors according to the first correlation and the first importance respectively corresponding to each initial characteristic vector and the preset qualification vector so as to obtain a first sorting queue.
In an embodiment, the transforming unit 73 is specifically configured to perform n × n operation on all the initial feature vectors to obtain a plurality of first transformed feature vectors, and use the first transformed feature vectors as target feature vectors, where n is a natural number and n is used to describe the n initial feature vectors.
In one embodiment, the customer qualification evaluation device 70 further comprises:
the operation unit is used for performing preset power operation on the initial characteristic to obtain a second conversion characteristic vector corresponding to the initial characteristic vector;
the transforming unit 73 is specifically configured to perform n × n operation on all the initial feature vectors to obtain a plurality of first transformed feature vectors, and use the first transformed feature vectors and the second transformed feature vectors as the target feature vectors.
In one embodiment, the transforming unit 73 includes:
the third sorting subunit is used for sorting the first conversion characteristic vector and the second conversion characteristic vector as conversion characteristic vectors according to a preset second sorting mode to obtain a second sorting queue;
a fourth obtaining subunit, configured to obtain a second preset number of the conversion feature vectors included in the second sorting queue, and use the second preset number of the conversion feature vectors as the target feature vectors.
It should be noted that, as can be clearly understood by those skilled in the art, the above-mentioned customer qualification evaluation apparatus based on combination characteristics and attention mechanism and the specific implementation process of each unit may refer to the corresponding description in the foregoing method embodiment, and for convenience and brevity of description, no further description is provided herein.
Meanwhile, the division and connection manner of each unit in the customer qualification evaluation device based on the combination characteristics and the attention mechanism are only used for illustration, in other embodiments, the customer qualification evaluation device based on the combination characteristics and the attention mechanism may be divided into different units as needed, or each unit in the customer qualification evaluation device based on the combination characteristics and the attention mechanism may adopt different connection sequences and manners to complete all or part of the functions of the customer qualification evaluation device based on the combination characteristics and the attention mechanism.
The above customer qualification evaluation apparatus based on the combination of features and attention mechanism may be implemented in the form of a computer program that can be run on a computer device as shown in fig. 8.
Referring to fig. 8, fig. 8 is a schematic block diagram of a computer device according to an embodiment of the present application. The computer device 500 may be a computer device such as a desktop computer or a server, or may be a component or part of another device.
Referring to fig. 8, the computer device 500 includes a processor 502, a memory, which may include a non-volatile storage medium 503 and an internal memory 504, which may also be a volatile storage medium, and a network interface 505 connected by a system bus 501.
The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032, when executed, causes the processor 502 to perform a method for assessing customer qualification based on combined features and attentiveness as described above.
The processor 502 is used to provide computing and control capabilities to support the operation of the overall computer device 500.
The internal memory 504 provides an environment for the operation of the computer program 5032 in the non-volatile storage medium 503, and when the computer program 5032 is executed by the processor 502, the processor 502 may perform a customer qualification assessment method based on the combination of features and attention.
The network interface 505 is used for network communication with other devices. Those skilled in the art will appreciate that the configuration shown in fig. 8 is a block diagram of only a portion of the configuration relevant to the present teachings and does not constitute a limitation on the computer device 500 to which the present teachings may be applied, and that a particular computer device 500 may include more or less components than those shown, or combine certain components, or have a different arrangement of components. For example, in some embodiments, the computer device may only include a memory and a processor, and in such embodiments, the structures and functions of the memory and the processor are consistent with those of the embodiment shown in fig. 8, and are not described herein again.
Wherein the processor 502 is configured to run the computer program 5032 stored in the memory to implement the following steps: acquiring client information corresponding to a client; acquiring an initial characteristic vector corresponding to the customer information according to the customer information; performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector; and inputting the initial characteristic vector and the target characteristic vector into a preset client qualification evaluation model based on an attention mechanism to evaluate the qualification of the client and obtain a client qualification evaluation result corresponding to the client.
In an embodiment, when the processor 502 implements the step of obtaining the initial feature vector corresponding to the customer information according to the customer information, the following steps are specifically implemented:
judging whether the customer information is numerical value information or not;
if the customer information is numerical value information, converting the customer information into corresponding initial characteristic vectors according to a preset numerical value conversion mode;
and if the customer information is non-numerical information, converting the customer information into the corresponding initial characteristic vector according to a preset word vector conversion mode.
In an embodiment, when the processor 502 implements the step of performing feature transformation on the initial feature vector to obtain the target feature vector corresponding to the initial feature vector, the following steps are specifically implemented:
acquiring a plurality of initial feature vectors corresponding to the customer information;
sequencing the initial characteristic vectors according to a preset first sequencing mode to obtain a first sequencing queue;
acquiring a first preset number of initial feature vectors contained in the first sequencing queue, and taking the first preset number of initial feature vectors as initial target feature vectors;
and performing characteristic transformation on the initial target characteristic vector to obtain a target characteristic vector corresponding to the initial target characteristic vector.
In an embodiment, when the processor 502 implements the step of sorting the plurality of initial feature vectors according to a preset first sorting manner to obtain a first sorting queue, the following steps are specifically implemented:
acquiring a preset qualification vector corresponding to the client qualification;
calculating a first correlation corresponding to each initial characteristic vector and each preset qualification vector;
calculating the first importance of each initial characteristic vector and the corresponding preset qualification vector;
and sequencing the plurality of initial characteristic vectors according to the first correlation and the first importance respectively corresponding to each initial characteristic vector and the preset qualification vector to obtain a first sequencing queue.
In an embodiment, when the processor 502 implements the step of performing feature transformation on the initial feature vector to obtain the target feature vector corresponding to the initial feature vector, the following steps are specifically implemented:
and performing n x n operation on all the initial feature vectors to obtain a plurality of first conversion feature vectors, and taking the first conversion feature vectors as target feature vectors, wherein n is a natural number and is used for describing the n initial feature vectors.
In an embodiment, the processor 502 further implements the following steps before implementing the step of taking the first converted feature vector as a target feature vector:
performing preset power square operation on the initial characteristic to obtain a second conversion characteristic vector corresponding to the initial characteristic vector;
when the processor 502 implements the step of using the first converted feature vector as a target feature vector, the following steps are specifically implemented:
and taking the first conversion feature vector and the second conversion feature vector as the target feature vector.
In an embodiment, when the processor 502 implements the steps of using the first converted feature vector and the second converted feature vector as the target feature vector, the following steps are implemented:
according to a preset second sorting mode, taking the first conversion characteristic vector and the second conversion characteristic vector as conversion characteristic vectors and sorting to obtain a second sorting queue;
and acquiring a second preset number of the conversion characteristic vectors contained in the second sorting queue, and taking the second preset number of the conversion characteristic vectors as the target characteristic vectors.
It should be understood that in the embodiment of the present Application, the Processor 502 may be a Central Processing Unit (CPU), and the Processor 502 may also be other general-purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, and the like. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
It will be understood by those skilled in the art that all or part of the processes in the method for implementing the above embodiments may be implemented by a computer program, and the computer program may be stored in a computer readable storage medium. The computer program is executed by at least one processor in the computer system to implement the flow steps of the embodiments of the method described above.
Accordingly, the present application also provides a computer-readable storage medium. The computer-readable storage medium may be a non-volatile computer-readable storage medium or a volatile computer-readable storage medium, the computer-readable storage medium storing a computer program that, when executed by a processor, causes the processor to perform the steps of:
a computer program product which, when run on a computer, causes the computer to perform the steps of the combined feature and attention based customer qualification assessment method described in the embodiments above.
The computer readable storage medium may be an internal storage unit of the aforementioned device, such as a hard disk or a memory of the device. The computer readable storage medium may also be an external storage device of the device, such as a plug-in hard disk, a Smart Media Card (SMC), a Secure Digital (SD) Card, a Flash memory Card (Flash Card), etc. provided on the device. Further, the computer-readable storage medium may also include both an internal storage unit and an external storage device of the apparatus.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, devices and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again.
The storage medium is an entity and non-transitory storage medium, and may be various entity storage media capable of storing computer programs, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a magnetic disk, or an optical disk.
Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present application.
In the several embodiments provided in the present application, it should be understood that the disclosed apparatus and method may be implemented in other ways. For example, the above-described apparatus embodiments are merely illustrative. For example, the division of each unit is only one logic function division, and there may be another division manner in actual implementation. For example, various elements or components may be combined or may be integrated into another system, or some features may be omitted, or not implemented.
The steps in the method of the embodiment of the application can be sequentially adjusted, combined and deleted according to actual needs. The units in the device of the embodiment of the application can be combined, divided and deleted according to actual needs. In addition, functional units in the embodiments of the present application may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a storage medium. Based on such understanding, the technical solution of the present application may be substantially implemented or contributed to by the prior art, or all or part of the technical solution may be embodied in a software product, which is stored in a storage medium and includes instructions for causing an electronic device (which may be a personal computer, a terminal, or a network device) to perform all or part of the steps of the method according to the embodiments of the present application.
The above description is only for the specific embodiments of the present application, but the scope of the present application is not limited thereto, and any person skilled in the art can easily conceive various equivalent modifications or substitutions within the technical scope of the present application, and these modifications or substitutions should be covered by the scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.
Claims (10)
1. A customer qualification evaluation method based on combined features and attention mechanism comprises the following steps:
acquiring client information corresponding to a client;
acquiring an initial characteristic vector corresponding to the customer information according to the customer information;
performing feature transformation on the initial feature vector to obtain a target feature vector corresponding to the initial feature vector;
and inputting the initial characteristic vector and the target characteristic vector into a preset client qualification evaluation model based on an attention mechanism to evaluate the qualification of the client and obtain a client qualification evaluation result corresponding to the client.
2. The method for evaluating customer qualification based on combined feature and attention mechanism according to claim 1, wherein the step of obtaining an initial feature vector corresponding to the customer information according to the customer information comprises:
judging whether the customer information is numerical value information or not;
if the customer information is numerical value information, converting the customer information into corresponding initial characteristic vectors according to a preset numerical value conversion mode;
and if the customer information is non-numerical information, converting the customer information into the corresponding initial characteristic vector according to a preset word vector conversion mode.
3. The method for evaluating customer qualification based on combined feature and attention mechanism according to claim 1, wherein the step of transforming the initial feature vector to obtain the target feature vector corresponding to the initial feature vector comprises:
acquiring a plurality of initial feature vectors corresponding to the customer information;
sequencing the initial characteristic vectors according to a preset first sequencing mode to obtain a first sequencing queue;
acquiring a first preset number of initial feature vectors contained in the first sequencing queue, and taking the first preset number of initial feature vectors as initial target feature vectors;
and performing characteristic transformation on the initial target characteristic vector to obtain a target characteristic vector corresponding to the initial target characteristic vector.
4. The method for customer qualification assessment based on combined feature and attention mechanism according to claim 3, wherein the step of sorting the plurality of initial feature vectors according to a predetermined first sorting manner to obtain a first sorting queue comprises:
acquiring a preset qualification vector corresponding to the client qualification;
calculating a first correlation corresponding to each initial characteristic vector and each preset qualification vector;
calculating the first importance of each initial characteristic vector and the corresponding preset qualification vector;
and sequencing the plurality of initial characteristic vectors according to the first correlation and the first importance respectively corresponding to each initial characteristic vector and the preset qualification vector to obtain a first sequencing queue.
5. The method for evaluating customer qualification based on combined feature and attention mechanism according to claim 1, wherein the step of transforming the initial feature vector to obtain the target feature vector corresponding to the initial feature vector comprises:
and performing n x n operation on all the initial feature vectors to obtain a plurality of first conversion feature vectors, and taking the first conversion feature vectors as target feature vectors, wherein n is a natural number and is used for describing the n initial feature vectors.
6. The method for assessing customer qualification based on combined feature and attention mechanisms according to claim 5, wherein the step of using the first transformed feature vector as a target feature vector is preceded by the steps of:
performing preset power square operation on the initial characteristic to obtain a second conversion characteristic vector corresponding to the initial characteristic vector;
the step of using the first converted feature vector as a target feature vector comprises:
and taking the first conversion feature vector and the second conversion feature vector as the target feature vector.
7. The method of claim 6, wherein the step of using the first transformed feature vector and the second transformed feature vector as the target feature vector comprises:
according to a preset second sorting mode, taking the first conversion characteristic vector and the second conversion characteristic vector as conversion characteristic vectors and sorting to obtain a second sorting queue;
and acquiring a second preset number of the conversion characteristic vectors contained in the second sorting queue, and taking the second preset number of the conversion characteristic vectors as the target characteristic vectors.
8. A customer qualification evaluation device based on combined features and attention mechanism, comprising:
the first acquisition unit is used for acquiring client information corresponding to a client;
the second acquisition unit is used for acquiring the initial characteristic vector corresponding to the customer information according to the customer information;
the transformation unit is used for carrying out characteristic transformation on the initial characteristic vector to obtain a target characteristic vector corresponding to the initial characteristic vector;
and the evaluation unit is used for inputting the initial characteristic vector and the target characteristic vector into a preset client qualification evaluation model based on an attention mechanism so as to evaluate the qualification of the client and obtain a client qualification evaluation result corresponding to the client.
9. A computer device, comprising a memory and a processor coupled to the memory; the memory is used for storing a computer program; the processor is adapted to run the computer program to perform the steps of the method according to any of claims 1-7.
10. A computer-readable storage medium, characterized in that the storage medium stores a computer program which, when being executed by a processor, realizes the steps of the method according to any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011605039.8A CN112686677A (en) | 2020-12-30 | 2020-12-30 | Customer qualification evaluation method and device based on combination characteristics and attention mechanism |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011605039.8A CN112686677A (en) | 2020-12-30 | 2020-12-30 | Customer qualification evaluation method and device based on combination characteristics and attention mechanism |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN112686677A true CN112686677A (en) | 2021-04-20 |
Family
ID=75454556
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011605039.8A Pending CN112686677A (en) | 2020-12-30 | 2020-12-30 | Customer qualification evaluation method and device based on combination characteristics and attention mechanism |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112686677A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113596270A (en) * | 2021-08-18 | 2021-11-02 | 中国平安财产保险股份有限公司 | Outbound strategy configuration method, device and equipment based on intelligent voice customer service |
-
2020
- 2020-12-30 CN CN202011605039.8A patent/CN112686677A/en active Pending
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113596270A (en) * | 2021-08-18 | 2021-11-02 | 中国平安财产保险股份有限公司 | Outbound strategy configuration method, device and equipment based on intelligent voice customer service |
| CN113596270B (en) * | 2021-08-18 | 2023-02-14 | 中国平安财产保险股份有限公司 | Outbound strategy configuration method, device and equipment based on intelligent voice customer service |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| WO2020107872A1 (en) | Company risk analyzing method, apparatus, computer device, and storage medium | |
| JP2021163471A (en) | Integrated clustering and outlier detection using optimization solver machine | |
| CN112541592A (en) | Federal learning method and device based on differential privacy and electronic equipment | |
| CN113240510B (en) | Abnormal user prediction method, device, equipment and storage medium | |
| CN112541575B (en) | Method and device for training graph neural network | |
| CN111680142A (en) | Automatic answering method and device based on text recognition and computer equipment | |
| CN111414533A (en) | Recommendation information generation method and device, electronic equipment and storage medium | |
| CN113095408A (en) | Risk determination method and device and server | |
| CN112712086A (en) | Data processing method, data processing device, computer equipment and storage medium | |
| CN112686677A (en) | Customer qualification evaluation method and device based on combination characteristics and attention mechanism | |
| CN107038165B (en) | Service parameter acquisition method and device | |
| CN111738356A (en) | Object feature generation method, device, equipment and storage medium for specific data | |
| CN111507461A (en) | Interpretability information determining method and device | |
| CN115983365A (en) | Model training method and device, computer equipment and storage medium | |
| CN109544348B (en) | Asset security screening method, device and computer readable storage medium | |
| CN113157987A (en) | Data preprocessing method for machine learning algorithm and related equipment | |
| CN109086207B (en) | Page response fault analysis method, computer readable storage medium and terminal device | |
| CN112330411A (en) | Group product recommendation method and device, computer equipment and storage medium | |
| CN112116381B (en) | Moon life prediction method based on LSTM neural network, storage medium and computer equipment | |
| CN112967134B (en) | Network training method, risk user identification method, device, equipment and medium | |
| CN113344613B (en) | Data matching method and device, electronic equipment and storage medium | |
| Armillotta | Two-stage weighted least squares estimator of multivariate discrete-valued observation-driven models | |
| CN116975701A (en) | User mining method, user mining device and electronic equipment | |
| CN117076977A (en) | Object classification method, device, computer equipment and storage medium | |
| CN117035416A (en) | Enterprise risk assessment method, enterprise risk assessment device, equipment and storage medium |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |